The Journal of Organic Chemistry
3178 KADIN C. 3- (2 ’-Pyridy1)-, 3- (3 ’-Pyridy1)-, and 3- (4’-Pyridy1)thioa period of 15 min a solution of 10 g (0.063 mol) of bromopyridine in 25 ml of dry ether was added to a well-stirred solution,of 0.063 mol of n-butyllithium in 32 ml of hexane diluted with 100 ml of dry ether under a nitrogen atmosphere. During the addition the temperature was maintained between -75 and -60’. After a second 15-min period of stirring, a solution of 6.4 g (0.063 mol) of 3-ketotetrahydrothiopheneI6 in 25 ml of dry ether was added in 15 min. The stirring of the gray suspension was continued for another 15min period raising the temperature to -40’. The reaction mixture was poured into 100 ml of water and the aqueous layer was extracted with ether (in the case of the 4-pyridyl compound chloroform must be used). The combined organic layers were dried (CaC19) and the solvent was evaporated. Removal of unused ketone through distillation at the water pump [up to 85’ (20 mm)] left 6.0-8.0 g (0.033-0.044 mol, 52-70%) of a viscous oil (for the 4-pyridyl compound a solid), mainly the tertiary carbinol. This oil was heated with 2.5 g of sulfur and 2.0 g of KHSO,. At 160’ water began to distil from the reaction mixture and hydrogen sulfide was evolved. The heating was continued for 30 min and the temperature was slowly raised to 225’. Steam distillation of the reaction mixture yielded crystalline plates of 3-(3‘-pyridyl)- and 3-(4’-pyridy1)thiophene (V and VI). With the 2-pyridyl compound an oil was obtained, which was further purified through preparative tlc on silica gel (Merck PF-254) with chloroform as eluent to remove some unchanged carbinol. Distillation furnished an analytical sample of 3-(2‘-pyridyl)-
phene (IV, V, and VI).-Over
r
Compd
MP, ‘C
%
C
IV V VI
27-28 75-76.5 138.5-139
10 10 2
67.16 66.70 66.99
Analysis, % Found N H
I I1
4.50 4.39 4.41
8.59 8.59 8.62
V
4.38
8.68
Pmr spectra Aromatic
1.7 (t-t) (I), 2.2-3.2 (m) (3) 1.7 (9) (2), 2.5-3.2 (m) (2) 1.7 (d) (2), 3.1 (d) (2) 1.7 ( s ) (2), 2.4-3.1 (m) (2)
Alkyl
7.2 (t) (2), 8.0-8.9 ( 4 (41, 9.0 (t) (3) 7.5 (t) (2), 8.2-8.9 (m) ( 4 ) ~9.1 (t) (3) 7.6 (t) (2), 8.1-8.8 (m) (4), 9.1 (t) (3) 7.2 (m) (l), 8.0-8.8 (m) (5)i 9.0 (t) (3)
8
19.71 19.94 19.75
Y C a l C d for CaHiNS-
67.12
TABLE IV Desulfurization product of
I11
TABLE I11 Overall yield,
thiophene (IV), bp 93-95’ (0.45 mm). See Table I11 for analytical data. Bromopyridine8.-The 2- and 3-bromopyridines were commercially available. The unstable1* 4-brompyridine was obtained from the commercially available hydrogen chloride salt through addition of an equimolar amount of concentrated sodium hydroxide solution a t 0’. Extraction with ether, drying of the organic layers (N&S04),and evaporation of the solvent below 30’ yielded almost pure 4-bromopyridine. Because of its instability, it was used immediately for further reactions. Desulfurizations.-The desulfurizations were accomplished by refluxing the individual highly purified pyridylthiophenes (I, 11, 111, and V) with 10 times their weight of Raney nickel W-517 in absolute ethanol for 30 min. The Raney nickel was removed by filtration through a glass filter and was washed carefully with absolute ethanol. The concentrated filtrates were dissolved in ether and dried (CaC12). Evaporation of the solvent yielded a crude oil giving pmr spectra (Table IV) consistent with butylpyridines (25-70%).
19.92
(15) F. A. Buiter, J. H. Sperna Weiland, and H. Wynberg, Rcc. Trau. Chirn., ES, 1160 (1964).
Registry No.-I, 3319-99-1; 11, 21298-53-3; 111, 21298-54-4; IV, 21298-55-5; V, 21308-81-6; VI, 21308-82-7; 2-phenylpyridine, 1008-89-5 ; 3-phenylpyridine, 1008-88-4; 4-phenylpyridine, 939-23-1. (16) J. P. Wibaut, J. Overhoff, and H. Geldof, {Ma., IC, 807 (1936). (17) H. R. Billica and H. Adkins, “Organic Syntheses,” Coll. Vol. 111, John Wiley & Sons, Inc., New York, N. Y., p 176.
An Unusually Facile Anilide Ethanolysis SAULB. KADIN Medical Research Laboratories, Chus. Pfizer & Co., Inc.,Croton, Connecticut 06340 Received April 7, 1069 Refluxing ethanol promptly converts the 2-methy1-1,3(2H14H>dioxoisoquino1ine-4-carboxanilides(Ia-e) into ethyl 2-methy1-1,3(2H14H>dioxoisoquino1ine-4-carboxy1ate(11) in good yields. The facility with which this ethanolysis occurs appears to be related to the higher energy state of I relative to 11. Lack of enol character of I, presumably due to steric hindrance, and the hydrogen-bonded stabilization of its enolate anion (V) impart substantial acidic character to this molecule, and this property provides the proton which is believed to catalyze the ethanolysis.
Quite unexpectedly, therefore, in the course of Treatment of carboxylic acid esters with amines is purifying 2’-chloro-2-methyl-l,3 (2H,4H)-dioxoisoquinogenerally a convenient method for the preparation of (Ib) it was discovered that reamides. Alcohols, which often facilitate this r e a c t i ~ n , ~ , lined-carboxanilide ~ fluxing ethanol converted this substance into ethyl are generally considered to be sufficiently poor nucleo2-methy1-ll3-(2H,4H)-dioxoisoquinoline-4- carboxylate philes as to allow the intermediate tetrahedral complex (11) in a yield of better than 90% and this unusual alcoformed by nucleophilic attack of an amine on carbonyl carbon to lead, irreversibly, to amide f ~ r m a t i o n . ~ holysis was subsequently found to be characteristic for related anilides (Table I). To gain insight into the Furthermore, the inertness of alcohols toward amides and anilides frequently suggests their use as recrystalmechanism of this uncommon and extraordinarily lization solvent^.^ facile reaction, the physical properties of these anilides and related substances were investigated. (1) R. B. Wagner and H. D. Zook, “Synthetic Organic Chemistry,” The 2 -methyl- 1,3(2H,4H) -dioxoisoquinoline-4- carJohn Wiley & Sons, Inc., New York, N. Y.. 1953, p 568. (2) R.L. Betts and L. P. Hammett, 3. Amer. Cham. Soc.. W, 1568 (1937). (3) R. Baltzly, I. M. Berger, and A. A. Rothstein, ibid.. 74, 4149 (1950). (4) J. F. Bunnett and G. T. Davis, ibid., 84, 665 (1960).
(5) 8. M. MoElvain, “The Characterization of Organic Compounds,” Rev. Ed., The Macmillan Co., New York, N. Y., 1945, pp 141, 189.
AN UNUSUALLY FACILE ANILIDEETHANOLYSIS 3179
"01. 34, No. 10, October 1969 TABLEI pKa' VALUES A N D ETHANOLYSIS RESULTS OF SUBSTITUTICD 2-METHYL-1 ,3 (2H,4H)-DIOXOISOQUINOLINE-~CARBOXANILIDES
R-
+
C2H50H Cr
is maximal. This contrasts with the nonplanar keto configuration of I which, because of the steric effect which would be engendered between the hydrogen on the anilide nitrogen and the hydrogen at position 5 of the isoquinolinedione moiety, cannot assume the low energy, resonance-stabilized enol configuration (VI) similar to that which occurs in I1 and I11 (VII).
0
I
*ZJ
0
NH~-Q +
-
R
0
0
Iv
111
I1 Compound
R
H 2-c1 IC 2-OCH3 Id 2-CH3 4-NOg Ie 0 See Experimental Section. Ia Ib
P K ~a'
5.68
4.38 5.49 5.36 insoluble
Yield of 11, ?7a
42 94
QR
QR OyN"
1
71 64 81
boxanilides are unexpectedly strong acids (Table I). Specifically, they are substantially more acidic than I1 (pKa' = 7.15) or 4-acetyl-2-methylisoquinoline-1,3(2H,4H)-dione (111) (pKa' = 6.60). Since a carbamoyl moiety is inferior to an acetyl or a carbethoxy function in its electron-withdrawing proper tie^,^,' it appeared that the reasons for the acidity of compounds of type I might be related directly to the facility with which they undergo ethanolysis. The nmr spectra of the carboxanilides are consistent with a keto rather than an enol configuration. For example, the spectrum of Ia exhibits a singlet at T 6.75 (N-methyl), a singlet at 4.5 (methine hydrogen), an aromatic multiplet, and a signal at -0.5 (amide hydrogen). That the assignments of the methine and amide hydrogens are as indicated was confirmed by the nmr spectrum of K,2-dimethyl-l,3(2H,4H)-dioxoisoquinoline-4-carboxanilide (IV) which is quite similar to that of Ia except that there is no signal downfield from the aromatic multiplet. Although the spectra were generally obtained, for solubility purposes, in deuterated dimethyl sulfoxide, a solvent which, in comparison with less polar solvents, is not expected to favor an intramolecularly hydrogen-bonded enolic configuration,8 a few compounds were sufficiently soluble to be examined in deuterated chloroform in which case the nmr spectra showed little change from those obtained in the former solvent. Contrary to those of the carboxanilides, the nmr spectra of I1 and I11 exhibit chemical shifts which are expected for compounds having strongly bonded enolic configumtions. The enolic proton of the former appears at T -5.9 (CCl,) and that of the latter at -6.8 (CC14). Evidently, part of the reason for the enhanced acidity of I compared with that of I1 and I11 is that ionization of I results in the formation of a planar, conjugated anion (V) in which a-orbital overlap (6) R. G. Pearson and R . L. Dilion, J . Amer. Chem. SOC..76, 2439 (19.53). (7) hl. Charton, J. Org. Chem.. 29, 1222 (1964). (8) M. T. Rogers and J. L. Burdett, Can. J. Chem., IS, 1516 (1965).
V
VI
0 VII, R = CH,, OC,H,
Therefore] in the case of I, but not of I1 and 111,ionization is probably enhanced by the formation of an anion which is configurationally more stable than its conjugate acid. When I1 was refluxed with aniline or o-chloroaniline in ethanol, the corresponding carboxanilides were detected (tlc) in trace amounts although the ester was recovered almost quantitatively. Thus, I1 is the favored product under equilibrating conditions despite the greater nucleophilicity of aniline compared with that of e t h a n 0 1 , ~lending ~ ~ ~ additional credence to the postulated energy differences between I and 11. As anticipated, conventional aminolysis of I1 in xylene proved to be a successful route to the synthesis of compounds of type I provided that ethanol was removed from the reaction mixture. A second feature contributing to the enhanced acidity of I is noted when the pKa' values of 23 anilides of structure I substituted in the ortho, meta, and para positions of the anilide moiety are plotted against the pK, values of the correspondingly substituted anilines." A straight line of slope 0.48 (correlation coefficient, (9) C. G . Swain, D . C . Dittmer, and L. E. Kaiser, J . Amer. Chem. SOC., 71, 3737 (1955). (10) Although Wepster and Verkade [ R e c . Trau. Chim., 61, 425 (1948)] showed that 2'-nitroacetanilide undergoes methoxide ion catalyzed methanolysis, we have found that 2'-chloroacetoacetanilide [H. E. Fierz-David and E. Ziegler, Helu. Chim. Acta. 11, 776 (1928)], a p-ketoanilide bearing some structural similarity to Ib, is recovered unchanged after refluxing in ethanol for 5 hr; no trace (tlo) of ethyl acetoacetate was detected. (11) L. P. Hammett, "Physical Organic Chemistry," McGraw-Hill Book Co., Inc., New York, N . Y . , 1940, Chapter 7.
3180
KADIK
The Journal of Organic Chemistry
0.97) is obtained. This indicates that the substituent effect upon the electron density about the aniline nitrogen is proportional to its effect upon the acidity of I, and that the latter effect is probably manifested by means of a hydrogen bond from the enolate anion oxygen to the hydrogen on the anilide nitrogen (V). A similar type of enolate anion stabilization has been proposed to explain the difference in acidity between N-salicoylmesitamide and salicylamide. l2 The fact that IV (pKa' = 7.57) is much less acidic than Ia-d also emphasizes the acid-enhancing effect of the hydrogen on the anilide nitrogen. I n conclusion, the unusual ease with which compounds of type I undergo ethanolysis seems to be related to the substantially higher energy state in which they exist relative to the reaction product. The lack of enol character of I, presumably due to steric hindrance, and the relatively greater stability of the planar, conjugated anion (V) resulting upon ionization of I impart substantial acidic character to this molecule, and this intrinsic acidity provides the proton which is probably required to catalyze the ethanolysis. l 3
that o-methoxyphenyl isocyanate was employed: mp 197-198' dec (acetonitrile). Anal. Calcd for CI~HI~NZOI: C, 66.66; H, 4.97; N, 8.64. Found: C, 66.69; H, 5.02; N, 8.71.
2-Methyl-l,3(2H,4H)-dioxoisoquinoline-4-carbox-o-toluidide
(Id).-A mixture of 4.4 g (0.025 mol) of 2-methylisoquinoline1,3(2H,4H)-dione and 2.6 g (0.026 mol) of triethylamine in 20 ml of dry dimethyl sulfoxide was treated, dropwise, with 3.5 g (0.026 mol) of o-tolyl isocyanate. The resulting solution was stirred a t room temperature for 3 hr and then poured into an ice-water mixture containing 5 ml of 6 AV hydrochloric acid. The precipitate which formed was filtered, dried, and recrystallized from acetonitrile to yield 5.1 g (73'%) of white crystals, mp 221-223' dec. Anal. Calcd for ClsHleNzOa: C, 70.11; H, 5.23; N, 9.09. Found: C, 70.45; H, 5.19; N, 9.17. 4'-Nitro-P-methyl-l,3(2H,4H)-dioxoisoquinoline-4-carboxanilide (le).-A solution of 4.94 g (0.02 mol) of ethyl 2-methyl-1,3(2H,4H)-dioxoisoquinoline-4-carboxylateand 3.0 g (0.022 mol) of p-nitroaniline in 100 ml of xylene was refluxed for 4.5 hr during which time solvent was slowly removed by means of a still head. A precipitate began to form as the mixtuw approaehed reflux temperature and became heavier nrith time. When the mixture had cooled, the precipitate was filtered and dried to yield 6.0 g (88'%), mp 235-236" dec. Attempted recrystallizations were unsuccessful. Anal. Calcd for C17H13N308: C, 60.17; H, 3.86; N, 12.39. Found: C, 60.45; H, 3.88; N, 12.74. 2-Methyl-l,3(2H,4H)-dioxoisoquinoline-4-carbox-N-methylExperimental Section anilide (IV).-A solution of 2.5 g (0.01 mol) of ethyl 2-methyl1,3(2H,4H)-dioxoisoquinoline-4-carboxylate and 1.1 g (0.01 Nmr spectra were recorded on a Varian A-60 spectrometer and mol) of N-methylaniline in 25 ml of xylene was refluxed for 1 are reported as T values relative to tetramethylsilane as an inhr during which time solvent was slowly removed by means ternal standard. Melting points are uncorrected. pK,' deof a still head. The reaction mixture was chilled in an ice bath terminations were performed in l :2 water-dioxane using a while 15 ml of hexane was added. The resulting precipitate was Metrohm Potentiograph. Isocyanates used were commercial filtered, dried, and recrystallized from benzene-hexane to yield materials. Each of the carboxanilides gave a positive ferric 1.8 g (58y0) of white crystals, mp 160-162'. chloride test. Anal. Calcd for C18HleN203: C, 70.11; H, 5.23; N, 9.09. 2-Methvlisoouinoline-1.3~2H.4H~-dione.~*-A solution of Found: C, 69.86; H, 5.22; N , 9.06. homophthalic acid in aqueous methylamine was evaporated to Ethyl 2-Methyl-1,3(2H,4H)-dioxoisoquinoline-4-carboxyfate dryness under reduced pressure and the residue was heated for (II).-A solution of 16.4 g (0.05 mol) of I b in 12.5 ml of ethanol 1hr in an oil bath at 180-190'. The melt was poured into ethanol was refluxed for 4 hr after which time ca. half of the ethanol was from which crystals, mp 122-124", were obtained (lit.16 mp 123'). allowed to evaporate. Cooling produced white needles which 2-Methyl-1,3(2H,4H)-dioxoisoquinoline-4-carboxanilide(Ia). were recrystallized from ethanol to yield 11.5 g (94yC)of ma-A refluxing solution of 35.0 g (0.2 mol) of Z-methylisoquinolineterial, mp 113-115' (positive ferric chloride test). 1,3(2H,4H)-dione and 21.2 g (0.21 mol) of triethylamine in 250 Anal. Calcd for CISHI~NO~: C, 63.15; H, 5.30; N, 5.67. ml of dry tetrahydrofuran was treated, dropwise, with a solution Found: C, 63.07; H, 5.33; N, 5.65. of 25.0 g (0.21 mol) of phenyl isocyanate in 10 ml of dry tetraEthanolyses of other compounds in Table I were effected by a hydrofuran. Refluxing was continued for 1.5 hr after addition similar procedure, using tlc to follow the course of the reaction. of the isocyanate was complete. The red-brown solution was A solution of 8.75 g (0.05 mol) of 2-methylisoquinoline-1,3poured into a stirred ice-water mixture containing 28 ml of (2H,4H)-dione in 80 ml of dry tetrahydrofuran was added to a hydrochloric aeid. The resulting precipitate was filtered, washed stirred suspension of 4.45 g of a 59.5% mineral oil dispersion of with water, dried, and recrystallized from acetonitrile to yield sodium hydride in 15 ml of dry tetrahydrofuran and the resulting 32.0 g (55y0) of white crystals, mp 243-244' dec. suspension was treated over 45 min with a solution of 6.0 g Anal. Calcd for C U H ~ ~ N Z O C,~ 69.37; : H , 4.80; N, 9.52. (0.055 mol) of ethyl chloroformate in 10 ml of the same solvent. Found: C, 69.52; H, 4.81; N, 9.51. A #an suspension developed during 19 hr of stirring at room 2 '-Chloro-2-methyl-l,3(2H,4H)-dioxoisoquinoline4-carbox- temperature after which time the reaction mixture was poured anilide (Ib).-Ib was prepared in 54% yield as described for Ia into an ice-water mixture containing 20 ml of 6 N hydrochloric exceDt that o-ehlorophenvl isocvanate was emuloyed: mp 212" acid. The resulting precipitate was filtered, dried, and recrystal213' dec (acetonitrife). lized three times from ethanol to yield 2.0 g (167c) of white Anal. Calcd for C17H13ClN20a: C, 62.10; H, 3.98; N, 8.52. crystals, mp 111-113' (positive ferric chloride test) whose inFound: C, 61.99; H, 3.97; N, 8.47. frared spectrum was superimposable on that of the material 2-Methyl-l,3(2H,4H)-dioxoisoquinoline-4-carbox-o-anisidide obtained by anilide ethanolysis. (IC).-IC was prepared in 42y0 yield as described for I a except 4-Acetyl-2-methylisoquinoline-1,3(2H,4H)-dione (III).-This compound was prepared by the method of Bailey and Swallow:l6 (12) R. M. Topping and D. E. Tutt, J . Chem. Soc., B , 1346 (1967). mp 115-117' mp 115'). (13) The possibility also exists that the ethanolysis proceeds through a I
.
I~
I
ketene intermediate similar to that recently proposed by Bruice and Holmquist [ J . Amer. Chem. SOC.,90, 7136 (1968)l for the hydrolysis of malonate and cyanoscetste esters. (14) S. Gabriel, Ber., 19, 2363 (1886). (15) -4. S. Bailey and D. L. Swallow. J . Chem. SOC.,2477 (1956).
Registry No.-Ia, 21389-75-3; Ib, 21389-76-4; IC, 21389-77-5; Id, 21389-78-6; Ie, 21389-79-7; 11, 21389-80-0; IV, 21389-81-1.
